Empowering Women in Tech: Harnessing AI for Gender Equity and the Future of Work with Sophia Matveeva

00:00 - Jo Shilton (Host)

on International Women's Day: 2024

00:16

Sophia is the CEO and founder of Tech for Non-Techies, an executive education and consulting company. She has contributed to the Harvard Business Review, Financial Times, the Guardian and Forbes on Entrepreneurship and Technology and hosts the top-rated Tech for Non-Techies podcast, so I'm sure I'm going to be safe in her hands today. She has also guest lectured at the University of Chicago, the London Business School and Oxford University. She is a startup mentor at the Chicago Booth Polsky Centre of Entrepreneurship and has advised leading accelerators, including Chicago Booth's new venture challenge and the Tech Stars Blackstone Launchpad. She holds an MBA from Chicago Booth and a BSc Honours in Politics from Bristol. She speaks English, Russian and French, and Sophia also sits on the advisory board to Riveter, which uses AI to predict customer trends for the world's biggest brands. So, Sophia Matveeva, welcome to Women WithAI.

01:09 - Sophia Matveeva (Guest)

Jo, I'm so happy to be here, and I'm really excited about the day that this is coming out, on International Women's Day and as the AI revolution is spreading. I think the conversation and the show that you are creating is going to be so important because, if we are not careful, the inequalities that are so prevalent in the tech industry, they're just going to be scaled with AI.

01:40 - Jo Shilton (Host)

Exactly, thank you. That's perfect because this podcast is dedicated to amplifying the voices and perspectives of women in the field of artificial intelligence. So, let's start with the basics. I'm sure our audience probably know a lot more about AI than I do, but as the host of tech for non-techies and you're speaking to a non-tech person what is AI?

02:01 - Sophia Matveeva (Guest)

So the way I like to think about it is it's basically machines doing things that we could have done before or that we could still do so really simple things. You could give a machine instructions to book you a table. This is literally what OpenTable software does. So I need a table for two people at 7 pm in central London, and then there is an algorithm which helps you make that reservation. So, previously, what did we do before OpenTable existed? We would pick up the phone; we would make a bunch of phone calls. Maybe we would have a secretary if we were really fancy to go and do that. So you see, it's a simple task that is automated that machines can now do. So that's simple AI. And then when we get into generative AI, it's basically when AI can make things, so can AI make new information. So, with this OpenTable, there is no new stuff that it's making. It's actually just completing a task for you, whereas generative AI is basically taking data that already exists and then changing it into something new and generating a new text, for example.

03:20 - Jo Shilton (Host)

Does that make sense? Yeah, totally. So it's to save time, really, isn't it? I think that was probably the first reason. Well, not the first reason, but that's the main thing that it does. It sort of frees up time.

03:32 - Sophia Matveeva (Guest)

Yeah, well, to save time, and for a business that means time with money, right. So to save time and to save money and also scale. So, in general, I like to think of technology as allowing us to save time, to save money, to be more efficient and to be more secure. So, generally, those are the four things that technology allows us to do. So not every single one, not every single technology, is going to be able to do all four. So, for example, you and I currently we are in different parts of London right now and people are going to hear this podcast all around the world.

04:11

So, essentially, our wisdom is being amplified by technology. So we're being scaled, and it's also being done in an efficient way, so we don't have to get a room together, get everybody flights to come and see us. People can listen to us. So we're scaling geographically, but also it's efficient because essentially, everybody can stay at home and carry on I don't know doing the dishes or whatever they're doing when they're listening to us. And so software, whether it's AI or whether it's just, you know, a normal algorithm, essentially it's scalability, efficiency, cost and security. Those are those are the things that it gives us and that's why businesses are so excited about it because essentially, it allows businesses, if used properly, to increase profits, which is, you know, as a businesswoman, something I'm always interested in. Yeah, definitely.

05:09 - Jo Shilton (Host)

So how do you use AI in your daily life?

05:14 - Sophia Matveeva (Guest)

You know I think, so I'll tell you how I use AI consciously, because all of us use.

05:19 - Jo Shilton (Host)

AI. Unconsciously, I don't think we realize how many devices we will have connected. I have a slightly unhealthy TikTok habit. How do you Siri? I use.

05:27 - Sophia Matveeva (Guest)

Alexa. So that's me using AI unconsciously because TikTok keeps on feeding me videos that keep me stuck to TikTok, as opposed to, you know, reading research papers on AI, which is what I should be doing, so that there's a lot of unconscious use of AI that we all that we all do, but some of the conscious things, especially for people who have any kind of content creation, ai is absolutely fantastic, but you've got to be careful here Because, for example, as I mentioned, I have the tech for non techies podcast and some of our interviews that video interviews, so they go onto YouTube. So previously, what we would do before AI is that my social media system would literally cut clips so find the most interesting bits, cut tiny clips and then put those tiny clips on TikTok or on Instagram or on LinkedIn for people to watch, and obviously that took my system time and that's time that I have to pay for. Now we have an AI tool called video AI, where we literally just put in the clip. It finds the bits that it somehow decides are interesting and then it clips them into Tiny reels, which then go up into social media. So what I want people to see here is that, yes, it's AI and yes, it definitely helps and it takes the content that we already have and it scales it.

06:52

But where I am not using AI is I am not using, say, chat GPT for To create content and then to pass it off as my own. And I think this is where the debate is is essentially, can you just get chat GPT, as you know, if somebody is a content creator or like, if you want to be a thought leader, like can you have really interesting, intelligent content created and then can you pitch it as an article to the Financial Times and get it published. So right now, that is not the case. So right now, I definitely do use chat GPT To get ideas to kind of debate, to like you know, sometimes you're staring at a blank piece of paper and you're thinking I've got to write something really smart and there's nothing happening in your head right now.

07:45 - Jo Shilton (Host)

I mean, and I think, sorry, and my friend get freezing, but I've caught up, now we can cut that bit out. That's how I use chat GPT as well. I use it to get ideas because it's like it's given you the basics that then you can work off. So that's how it saves time for me. I was just checking that you could hear me. Oh, I hadn't frozen, sorry.

08:08 - Sophia Matveeva (Guest)

On Riverside. Sometimes there is a delay just because of all the things that actually the AI is doing in the background. But you know one of my clients, she actually is a professor of AI at Texas A&M University, which is one of the top universities in the US and the top in Texas, and you know the Texans think that they're basically Well, that they're a law onto themselves, anyway. So she's a professor of AI that I actually interviewed on my podcast and she says that she sees chat GPT as her intern. So clever, very keen, you know, that kind of like kind of Hermione from Harry Potter, really keen, has read everything but has no actual life experience and so Really sweet.

08:58

But you probably shouldn't pass Hermione's work off as your own, at least when she's a 13 year old. So yes, so for research, take what Hermione is giving you, but then as a senior person with experience, you know nuance, you know what's really happening in the industry. But it's really helpful to have something to essentially Critique it, because often a lot of chat GPT something and then it'll give me an answer nothing or this is wrong, this is completely incomplete, this is, this is missing a cool, crucial point. And so then I basically start writing something original and interesting as a response to the incomplete thing that yeah, completely.

09:42 - Jo Shilton (Host)

So I think AI is making us more creative in a way. Would you agree? That is that how you see it.

09:50 - Sophia Matveeva (Guest)

It's making us so if you use it properly. Because, if you like, I remember we had when it first came out, and you know my social media system he's quite young, very keen, so kind of kind of like Hermione, very keen, quite young, not yet hugely experienced and so I thought, okay, well, I'm writing all of these newsletters, you know, as you said, I've written for the Harvard Business Review, so how about we take the content that I've created and essentially ask chat GPT to create some social media posts Based on what I've said? And so we tried that and it was a disaster, because I was like I have original thought there and literally all originality has just been squeezed out and it's just become, you know, generic, inspirational business stuff which is really really boring. And if we had just posted that, like I would have been thoroughly ashamed Because you can tell.

10:49

But essentially we looked at it and and we're like, okay, no, this is not enough, but we know where it's not enough and so we could change it. And I think AI makes you more creative when you respond to it, when you decide that, okay, this is not enough. This is where I could, this is where my human creativity Is going to matter, whereas I think some people mistakenly think, oh, I'm just gonna, you know, ask it something and then I'm gonna post it and then good things will happen.

11:25 - Jo Shilton (Host)

And most of the time that's not going to work and most of the time people can, and especially being based in the UK, it always has an American spelling or an American voice, even if you tell it that you're in the UK, or that it's for For that audience and that's the Well, not the problem.

11:41

But the issue is that it doesn't know the nuances of your own audience and what you want to say. So I mean, do you see that it's taking so potentially, it's freeing up time and it's handling those junior roles and those junior tasks? I mean, do you think that is a special significance for women In this industry? Is it those roles that women would normally take that are now maybe at risk? If it's the junior interns, I mean, you said your social media intern was a male, so clearly not. It was probably going to affect everyone. But yeah, what do you think the risks are?

12:17 - Sophia Matveeva (Guest)

Well. So there are risks unless you take the future into your own hands, basically. So I can literally talk about my social media assistant. He has become more productive because he has learned how to use AI tools, because there's a whole bunch of stuff I need him to do Like. If I possibly could get him to be awake and working for me 24 seven, that would be amazing. But you know, apparently he has human rights and needs to sleep.

12:48 - Jo Shilton (Host)

so we're not allowed to do that, Not yet anyway.

12:55 - Sophia Matveeva (Guest)

And so you know, my to-do list for him was as long as my art. But with AI, that means that, for example, he's not sitting there and trimming videos. He gets the AI to trim the videos and you know, about 60% of the results that he gets are good. So he throws out the other 40 and then he uses AI to find the right times to post things, which means that he can then be creative. And then I can say to him okay, you know, I am looking. So one of the things that we're now doing is we're working more and more with law firms. So when law firms want to work with tech clients, that's a really perfect client for us, because they're a non-technical audience. They don't need to become a technical audience, but they need to understand technology so they can serve their clients better. And so I can say to him look, this is the audience that I need. Can you go and find, you know, places where I can write an op-ed or podcast that law is listened to, and so on, and so that means that he is more productive because he's learned to use these tools. However, if people that you know, if I wasn't his boss and if I didn't have, essentially, knowledge of AI and knowledge of creative, and also, if I wasn't entrepreneurial, then I would be like, well, okay, well, you can't do this, so this job was eliminated.

14:15

So there's a really great phrase that I heard from a Harvard Business School professor, karim Lakhani, who said that it's not AI that's going to take jobs. It's humans who know how to use AI who are going to take jobs. And I think this is it. And like we're now at this point where we can decide. It's up to us. We can decide Is AI going to make me more productive? Because can I then go and say to an employer look, I can do all of these tasks and I can do all of these tasks basically with the same amount of time. Like you're hiring one person who can really do the task of three, so you can do that. Or you can think, okay, well, my job is being eliminated, that's it.

15:05 - Jo Shilton (Host)

It's spotting those opportunities completely. And you know, this is International Women's Day, so I think that we need to be looking at AI as an empowerment tool for women, and so I think we need to encourage more women to get into this field to learn how to use it. And how do you see the best ways to do that?

15:27 - Sophia Matveeva (Guest)

So, unfortunately, have you heard that term office housework? So it's like, you know, when you're in an office and there's just like at somebody's birthday and then a cake needs to be organised, or like somebody's leaving and then a card needs to be bought and a gift, so usually it's the women that are tasked with these things. So I mean it would make sense that you know the office manager or the secretary is going to do that, because that's kind of you know, if it's a team secretary and somebody in the team is leaving, that makes sense. But even when it's actually not, you know, let's say that there are male partners and female partners in a marketing firm. Generally, this kind of office housework, this, this is a luxury falls onto women and you know there are all sorts of reasons why that happens and it's crap and it happens.

16:23

And so you know, ladies, when you're listening to this, think about all the kids that bought all those cards that you bought, all that and like remembering, like remembering when it's somebody's birthday, babies, all of that stuff, all of that brain capacity, right. And you are not going to get promoted and you're not going to get paid more because you went to M&S and bought the cake yeah, totally agree. And yet so you're not going to get promoted for you, you're not going to get paid more.

16:55 - Jo Shilton (Host)

You're out of pocket and yet you're doing it.

16:58 - Sophia Matveeva (Guest)

Yeah, yeah and like it, it sucks, right. And so women in general are tasked, are tasked with these. You know drudge, bits of bits of office drudgery, and we can't use AI to do everything, but AI is going to allow us to eliminate at least some of that drudgery. I mean, we still can't have, like, robot cake deliveries.

17:21 - Jo Shilton (Host)

Well, you can in some places.

17:24 - Sophia Matveeva (Guest)

Actually, yes, you can. So at least some of that drudgery, and because women in particular tend to be given more of the admin, even when they are super senior. I think the way we fight that is you know, we fight that by saying no, why am I doing this? And I've literally. I remember I was working at a company and we were in a meeting room and there was a guy who was more junior than me and I was asked to go and get the printouts and give them to everybody and I said, well, why doesn't, why doesn't he go do it? And literally I was told oh you know, sophia, don't go, don't make a fuss, yeah and awful, and I literally just I mean I can feel that.

18:10 - Jo Shilton (Host)

I know the feel anger inside, just on your behalf, that that even happened and it but it does.

18:15 - Sophia Matveeva (Guest)

Yeah, but you know it's this, this kind of like everyday sexism, happens to all of us, right?

18:24

And so, on one hand, we do need to try to stand up for ourselves when we can, but also, frankly, sometimes it's not politically expedient, and this is I hate saying that, but sometimes it's true. So we need, we need to stand up when we can, but other times when actually you know what, like I'm not going to go and spend my political capital on fighting this. I can just get AI to do whatever this nonsense is, so I can spend the rest of my time actually doing strategic stuff, strategic work that wins their account and that gets me paid more. So I think, you know, I'd love to say that, yes, let's all go and, you know, fight, but sometimes, like it's, it's, it's unrealistic to say that we can fight every single injustice and also, still, you know, get on the account that we want, get the client that we want and get the promotions that we want. And you know, as an aside, there is this audio book that I listened to which I really loved. It's called Machiavelli for women.

19:34 - Jo Shilton (Host)

We can like how. I'll put a link to that in the show notes.

19:39 - Sophia Matveeva (Guest)

It's nothing to do with AI, but it's just. You know the reality of okay, what's it like to be a woman in the works, but you can't. You can't openly fight every injustice, but you can be smart about it.

19:51 - Jo Shilton (Host)

Yeah, and I think I think that's because women are great at multitasking, and I think maybe that's one of the things that you can use AI for as well, because it can be doing lots of different things or you're using lots of different tools to do things. Do you think AI? Do you see it as being more female or male, or does that not matter? Did you lose me?

20:17 - Sophia Matveeva (Guest)

We had a bit of a luck. Could you say that again?

20:21 - Jo Shilton (Host)

Yeah, I said that women are great at multitasking and we can do lots of things at once, and I think that's what AI is good for, because you can have either lots of different tools running, or you're using it to do lots of things, or it's freeing up time so you can do more things. Do you think AI is viewed as being female or male, or does it not matter? Or is it who's interacting?

20:47 - Sophia Matveeva (Guest)

So it's interesting because a lot of the time, AI tools are created to be female. So if you look at Alexa, Alexa, Siri, all of them they have female voices, which is interesting because we are socialised to believe that giving commands to a woman is fine, Giving out how orders to a woman is okay, Whereas if it was a man, like oh no, we can't give orders to a man. I actually do think that there are some pretty dark things in gender socialisation in terms of AI which we need to question, especially, I would say, for those who have children, Because we don't notice it. I didn't notice the whole that AI voices tend to be female until somebody said that to me and I was like, oh, that's really true. And I would say that to people who have children when you have, say, Alexa at your house, you are unconsciously programming kids to think that barking orders at a woman is fine, and I mean we all kind of fall for this socialisation. So I mean I certainly do too. So I would just say just try to be as conscious and aware of it as possible.

22:11

And in terms of, I think what women can do to participate in the conversation is A well, just use the thing use as much as possible, but also, if something is rubbish and it doesn't work, then question why it doesn't. So if, for example, you work in an organisation that maybe makes an AI tool, you know, going back to law firms, lots of law firms, especially the big law firms you know Magic Circle, the huge international law firms they are using all of their case history to make their own AI tools and so you can actually influence. You can actually say like, okay, I'm getting really terrible results in this. Why is that? And I think that confidence to just you know to say and to speak up and to say this isn't working for me. This needs to be fixed.

23:12

This is the one takeaway that I would want women to have from this conversation is that these are not fixed things. A lot of the time they are created by men who don't have the problems that we do, who don't face the issues that we do, and so often they would programme things into the AI that we had that you know that definitely don't work for us, but they didn't do it maliciously, they just don't know. And it's like do you know the story of when the Apple Watch was first released?

23:46 - Jo Shilton (Host)

No, I'm wearing one now. So now I feel bad that I don't know.

23:51 - Sophia Matveeva (Guest)

Oh no, so this was ages ago. This was so. I think I did an episode on it, I don't remember. Anyway, so this was ages ago when it was first released. So you know, all these Apple engineers who you know, who had been a fortune, got up to unveil the Apple Watch, and it was a bunch of guys and they were saying, oh, you know, it tracks all of your metrics and even you know your heart rate, how much you've walked, and even tracks how much beer you drank. It's like, oh yay, great. And then somebody in the audience happened to be female said and does it track the menstrual cycle? And they were like, oh no, didn't, didn't think of that. It tracks everything. It tracks everything apart from this thing that happens every month to 50% of the population.

24:37 - Jo Shilton (Host)

I know, ridiculous, because I do use it for that as well. But the hilarious someone at work. He said he got his Apple Watch and he said, oh, it does cycle tracking and he literally thought it was to sack his bicycle rides into work. I was like, no, it's not that kind of cycle. Yeah, you don't. You don't have one.

24:57

don't even know if it was the: 1990

25:23 - Sophia Matveeva (Guest)

Yeah, and as a result, more women have died in car crashes, and so what I would say to women is that this happens at very large Companies, which hire Apparently, the best people in the world and pay them a ton of cash and they make these Excuse me, but kindergarten blunders. You know they make these idiotic mistakes and they are still making them, and so what I would say is that if you are using a tool and you just see some obvious stupidity, don't dismiss it. I think we, as women, we often tend to think oh you know, I'm pretty good I don't speak up enough.

26:07

Oh, I didn't use it properly, yeah, like it's, it's my fault. We think, oh, I, there's something that I did that I don't understand and it's my fault. Well, actually, a lot of the time it isn't a lot of the time. It just wasn't made for you and it should be, because, if you're using it, that's why we need our seats at the table, because we do make up 50% of the population.

26:34 - Jo Shilton (Host)

So Well, yeah, well, now I mean, because you've built AI, haven't you? And so you don't even need to be a completely techie person. So can you, can you tell?

26:45 - Sophia Matveeva (Guest)

us about that so.

26:48

You know it's interesting. So when, if I had known that I was going to build an AI algorithm Before I started doing it, I would have freaked out, but I kind of didn't know that this is what I was doing. I mean, I only realized what I was doing when it was already happening. So I'll tell you what happened. So my first tech company was a retail tech business, so we had a consumer app where women would take photos of what they wanted to wear or buy and ask Professional stylists for their feedback. Like should I wear? Should I wear this skirt with this dress, or should I buy this, even though you know it's the wrong color, but it's 80% off? Oh, we're tempting. And so, yeah, oh, it's lime green and it's two sizes too small, but it's Dolce Gabbana and it's 80% off. I mean, we've all been there, right and so. So there was the consumer bit, and then that would allow us to see what people were were buying or not buying, and then sell that data analyzed to retailers and to brands and we needed to create, we needed to decide how people who would download their app. You know you could also see what other people were wearing or what other people buying. You know, there was a content feed and Some people had friends in the app, but it wasn't an Instagram type thing, it was like most, most people didn't know anybody else in the app, but some people did, and so we kept on trying to figure out Okay, how do we serve content to people in a way that they find interesting, but is? But also isn't just showing them exactly the same people, because that's boring, and and so I remember I was having a conversation with my CTO and we're like well, okay, so what do we prioritize that people see? Do we prioritize that people see their friends? Do we prioritize that people see the people who are voted the most stylish, consistently? Do we prioritize that people see the new users? Because, you know, like fresh blood. And we were having this discussion and he said something like well, you know, why don't we just vary it? And so we literally got a piece of paper and we're like okay, so if we have friend, followed by new person, followed by like most stylish, like what do? And then we kind of just have this variation, just like this like, and so the most important people are friends and then the least important people are kind of the most stylish. Like well, we don't know, let's use this logic, let's release it and see what happens. And Then, as I realized later, he was like okay, so this is basically the new content algorithm prototype, let's release this prototype and a beta sit. And I was like and you see what? I just the thing that I just explained to you.

29:50

Everybody could follow it right, and you could also see that You're like I didn't know the answer. There is no right answer, because you know, some of our friends are really stylish and I want to see what they're wearing. Other people. I'm like I love you, but you know, whatever I'm not wearing. No, yeah, I love you, but not because of you know what you wear, and and so you make these assumptions.

30:17

And then what an engineer does is they, they work with you to help you understand the assumptions that you're making. So you can really kind of put those assumptions basically into a bunch of mathematical equations and then those mathematical equations become code and then you know, you serve the algorithm, you serve that content, and then you see, okay, a people are engaging more or a people engaging less. And then you see, okay, people are really interested in this kind of thing. They're really not interested in this kind of thing and then you alter it and I think you know this is a really great example for A podcast that's going to be listened to primarily by women, because Nothing that I just said is particularly technical and nothing that I just said is something that you know you can't do, because it's just logic, like if you are capable of you know good, having a job You're capable of logic, and so then you're capable of, and you're capable of questioning the algorithm and you're capable of essentially saying what data has gone into making this algorithm.

31:29

Are there, is that equal amount of data of men and women, for example? You could ask, okay, is is this algorithm? We're producing past data. Because this is important, because right now in our world we live in a very unequal society. You know, in the UK Female founders only get 2% of Finch capital funding. The gender pay gap is enormous. So, like we right now Live in a society that is getting better, but phenomenally, slowly.

32:08

So if we have an algorithm that basically takes what we have and says oh, let's carry on doing that.

32:14 - Jo Shilton (Host)

I'm not on board. We need to make the change. I don't.

32:18 - Sophia Matveeva (Guest)

Yeah. And so we can literally say is this algorithm using past data and just extrapolating it to the future? And then, if they say yes, you can think well, is this what we want, or do we essentially want to program the future that we want?

32:35 - Jo Shilton (Host)

Because sometimes we do so, there are many opportunities for women. You've made me really excited about this journey that I'm starting as well, and hopefully our audience is going to come with us just to learn more about AI and how we can empower women to use it and how it impacts on us and how we can be part of changing it and making it better for all of us.

32:58 - Sophia Matveeva (Guest)

Yeah, absolutely. Well, I mean, the main thing is just don't put anybody on a pedestal. I think we have this tendency to put you know somebody who's like an AI engineer onto a pedestal and they're probably really good at what they do, but they don't have your perspective and they need your perspective. So you know, they might be as terrified of you as you are of them. So offer them your perspective. You know, kindly and politely, but offer them your perspective because they'll probably benefit from it.

33:28 - Jo Shilton (Host)

Yeah they may be the experts in building it, but they need the input and it's only ever going to be as good as the people that have input the data or input the thoughts, and so where it's going Exactly. Thank you, sophia. I mean, where can our audience learn more about this? Have you got any recommendations for podcasts or books or people to listen to?

33:49 - Sophia Matveeva (Guest)

So obviously the Tech for the Techies podcast.

33:54

Yeah, listen to the Tech for the Techies podcast and follow me on LinkedIn. So I put up lots of stuff on AI, but also on technology in general, on there. And there is a. There is a book called competing in the age of AI. Hang on, I'm just going to make sure that I get the proper, the full name of it. It's competing in the age of AI. Yeah, competing in the age of AI strategy and leadership when algorithms and networks run the world. So it's it's quite a thick read, so you're not going to want to have it on the beach and I actually I interviewed one of the authors and I did a book summary on my podcast.

34:44

If you can't, we bother to read the whole book, but the book is really really good because it's written by two Harvard Business School professors and it's about AI, but from a business perspective. So, yes, it teaches you basically what algorithms are, but it's not a computer science text. So, yes, like you know, I've read it with a highlighter or lots of post-its, but definitely, as a non-technical person with a business background, you will be able to understand it.

35:12 - Jo Shilton (Host)

That's fantastic, thank you, and you mentioned people can find you on LinkedIn, so we'll put some links up and all your tags in the show notes, but thank you for coming on Women with AI. I hope you've enjoyed it as much as I have. I've learned a lot, so thank you.

35:29 - Sophia Matveeva (Guest)

I loved it and everybody listened to this. Happy International Women's Day.